|
The trm Report - November 2007 Actuarial advice can take trustees and employers only so far. Access to knowledge about other pension schemes takes them further. If you want to know the value of your house, you expect the valuer to compare your property with similar ones in the area. No one with their wits about them would trust a surveyor who ignored the market. Likewise, no one values a business – or trade marks, film rights, even works of art – without looking at benchmarks. In short, pretty much everything gets valued by reference to comparables. Except pension schemes, that is. Is there a reason why pension schemes should be any different? Maybe actuaries just don’t need to compare their valuations with others. But the Treasury’s recent Review of the Actuarial Profession came to a rather different conclusion. The report’s author, Sir Derek Morris, was of the opinion that “it is essential that users regularly review the performance of all their actuarial advisers”. One of the reasons he gave for regular market testing of actuarial advice was to pick up “concerns… about the actuary’s willingness to discuss assumptions, methodology or risk parameters.” Now that the Pensions Act 2004 requires trustees to take the ultimate legal responsibility for valuation assumptions and the funding plan, wouldn’t it be reassuring for trustees if they knew how their valuations compared with others? Couple that with the statutory requirement that schemes must have sufficient assets to cover an agreed measure of their liabilities, how can any trustees not want to benchmark their valuation, before signing off? And wouldn’t members think it was in their best interests that the trustees had that information? So why aren’t comparisons a routine feature of the pensions scene? One argument I have heard used is that comparisons don’t fit well with a “scheme specific” funding regime. Try telling that to valuers of businesses or works of art. Any valuer who knows what he or she is doing will be “specific” about the subject matter of the valuation. But that doesn’t mean ignoring the market. Quite the reverse. It means finding your place in the market. And the valuer can do that only if they know how the market is behaving. The problem in the past has undoubtedly been the lack of data. A few surveys are published from time to time, but too little, too late. By the very nature of the surveys – published after the valuations have been finalised – they don’t provide current data for schemes being valued. Better still would be a live comparison able to access information that told you the key assumptions that schemes were using currently and across a range of actuarial advisers: rates of return, longevity, inflation and salary escalation. In fact, why stop there? With a database of assumptions, the obvious next step is to combine the key assumptions into an index showing the overall strength of the actuarial basis. After all, the interaction of assumptions is notoriously difficult to assess. And funding levels also. It would be helpful to know how well-funded schemes are – and how their actuarial liabilities are measured – and to know all this before finalising the valuations. When I realised, last year, that no one had built such a database, I decided to design one myself. The primary goal was that the database should shine a light on an area which has previously been shrouded in mystery. 1. The database absolutely had to be current. 2. It had to be totally confidential. 3. It had to fit with scheme specific funding. To make the database current, schemes should start using it before they had finalised their valuations. Trustees and employers should be able to start feeding in assumptions and valuation results they were considering using. By exchanging information before a final decision is made, they could engage in a confidential online information exchange with other schemes. As each scheme re-thinks the assumptions it wants to use, the revised information can be submitted. This would serve two purposes: it would enable the scheme to see how the revised assumptions and funding levels compared with other schemes and, at the same time, it would automatically update the database with the latest thinking. But it also needs to support a scheme-specific funding regime. Simply comparing your scheme with a central set of assumptions would fly in the face of the current regulatory impetus – and wouldn’t help trustees or employers in their discussions with each other. Addressing this required some subtlety. But it soon became apparent that three features would enable this to be achieved. First, the comparison needs to show the range of other schemes’ valuations. So, in addition to the central figures, the comparison charts must show the range of conservatism from, say, the bottom ten per cent of schemes in the database to the top ten per cent. This probably sounds more complicated than it is in practice, so long as the software is made user friendly and completely confidential. Third, and last, since scheme specific funding is the result of a negotiation between trustees and employers, there would be real advantages in the database keeping separate information on the different funding proposals submitted by trustees and employers. Keeping separate data is easy to do. Ensuring that trustees would not have access to the employer’s input, and vice versa, was also an important but straightforward component. More complex, but potentially very revealing, would be to devise, using standard data, measures which compared and contrasted the potentially different perspectives of employers and trustees.
So, on the one hand, trustees clearly have an interest in minimising the extent to which the scheme’s funds fall short of the amount needed to pay benefits if the employer were to fail. On the other hand, the employer cannot ignore the portion of shareholders funds which will have to be invested if the funding target is to be met. By capturing this data separately from employers and trustees and aggregating it across groups of pension schemes so that the confidentiality was totally preserved, it is possible to produce reports in a format which enable one user’s scheme to be compared with many others. The system, known as CompAct Analysis, now exists via the internet. All it needs is for schemes to populate it with data. Through CompAct Analysis, pension scheme trustees can assess the strength of the proposed valuation assumptions and funding levels. The comparison is conducted before the valuations have been finalised, so the information can be used as a decision tool for valuations required by the Pensions Act 2004. CompAct Analysis is completely confidential. Many in the pensions world now believe that comparing assumptions is a critical part of good scheme governance. Even if your scheme has completed a valuation under the 2004 Act regime and submitted the results to the regulator, it is not too late to submit the information to CompAct Analysis to help build up the database. So what can we expect to happen if schemes begin to compare valuations? Many schemes now set the investment return assumption by reference to the scheme’s investment strategy. So a scheme with an investment portfolio that is more than averagely cautious would want to set the assumption in a way that reflects that decision – perhaps by comparing its own assumption with schemes that have a similar investment strategy. The current regulatory regime places an emphasis on the strength of the employer. So it would be natural for trustees to treat that as a factor that guides the comparisons they carry out. A sophisticated comparison service should enable that to happen. Simon Carne |
 |
|
|||||||||
|
|||||||||||
